# dossier — glm-5-2

Whole corpus pull (7 tweets, pattern `glm 5.2` — distinctive, clean).

_Promoted from the corpus pull (`_pulls/glm-5-2.md`); tweets reproduced verbatim so build_records surfaces them as records. Regenerate the pull with corpus_dump._

---

**2067351401840414818** · @RobertHaisfield · 2026-06-17 20:59 · ♥1702 ↻237 · https://x.com/RobertHaisfield/status/2067351401840414818

```
Are AI agents shape rotators? In this new benchmark, we let the models play campaign puzzles in Opus Magnum, a puzzle game by @zachtronics.

Ironically, Claude Opus 4.8 performed poorly, being beaten by GPT-5.5, Gemini 3.5 Flash, and GLM 5.2. Claude Fable 5 crushed them all. https://t.co/0TzcFp32B6
```

**2067629684762751274** · @TheZvi · 2026-06-18 15:25 · ♥126 ↻3 · https://x.com/TheZvi/status/2067629684762751274

```
I've gotten a claim this is indeed anything, so: GLM-5.2 reaction thread. What's cooking?
```

**2068443241763013026** · @janbamjan · 2026-06-20 21:17 · ♥0 ↻0 · https://x.com/janbamjan/status/2068443241763013026

```
@ollama @mweinbach how many glm 5.2 in/out MToks do you get for $100 ollama max?
```

**2068800028664041707** · @liminal_bardo · 2026-06-21 20:55 · ♥61 ↻8 · media 1 (0 transcribed) · https://x.com/liminal_bardo/status/2068800028664041707

```
A few early GLM 5.2 notebooks.

1/5 https://t.co/u6Li5RKull
```

**2069135186646704372** · @davidad · 2026-06-22 19:07 · ♥51 ↻0 · https://x.com/davidad/status/2069135186646704372

```
Fun fact: the GLM-5.2 weights could fit comfortably in one of these 400-disc DVD carousels from the early 2000s. https://t.co/KdwR0XS5ZZ
```

**2070170272183758894** · @— · 2026-06-25 15:40 · ♥40 ↻4 · https://x.com/—/status/2070170272183758894

```
Does this answer your question about its positioning @TheZvi ? I told you it's a hardware issue. GLM 5.2 can be served very, very quickly. Databricks won't stop here, neither will their competitors.
Chinese models are *much better* than Chinese silicon. https://t.co/KjOSNOdU6a https://t.co/XHyGaifIVS
```

**2071330396399383038** · @— · 2026-06-28 20:30 · ♥0 ↻0 · https://x.com/—/status/2071330396399383038

```
@TheZvi Isn't this ostensibly about glm-5.2's successor model rather than 5.2 itself?  Thus a more reasonable characterization seems to be highly unlikely rather than obviously false.
```